The number of large data centres operated by hyperscale companies recently passed the 1,000 milestone – and new data from Synergy Research Group shows that they now account for 41% of the worldwide capacity of all data centres.
Just over half of that hyperscale capacity is now in own-built, owned data centres with the balance being in leased facilities. With non-hyperscale colocation capacity accounting for another 22% of capacity, that leaves on-premise data centres with just 37% of the total.
This is in stark contrast to six years ago when almost 60% of data centre capacity was in on-premise facilities. Looking ahead to 2029, hyperscale operators will account for over 60% of all capacity, while on-premise will drop to just 20%.
Over that period, the total capacity of all data centres will continue to rise rapidly, driven primarily by hyperscale capacity growing almost threefold over the next six years. While the colocation share of total capacity will slowly decrease, colocation capacity will continue to rise steadily. On-premise share of the total will drop by almost 3% per year, though the actual capacity of on-premise data centres will remain relatively stable.
Data centre capacity trends
The Synergy data is based on a combination of several detailed quarterly tracking research services, which enables it to build a comprehensive analysis of data centre capacity, with breakouts by region, country and metro markets.
The hyperscale research is based on an analysis of the data centre footprint and operations of the world’s major cloud and internet service firms, including the largest operators in SaaS, IaaS, PaaS, search, social networking, e-commerce and gaming.
The colocation research is based on Synergy’s in-depth tracking of the colocation market, including quarterly data on over 290 individual companies. The enterprise on-premise analysis is based on Synergy’s tracking of the data centre hardware market.
“The mix of data centre capacity is quite different region by region, an example being that hyperscale owned data centre capacity is much more prevalent in the US than in either Europe or the APAC region,” said John Dinsdale, a Chief Analyst at Synergy Research Group. “However, overall, the trends are all heading in the same direction – and it is easy to see what is behind these trends.
“In 2012, enterprises spent 12x as much on their data centre hardware and software as they did on cloud infrastructure services, while today they spend 3x more on cloud services than they do on their own data centre infrastructure.
“Add to that the huge growth in SaaS and consumer-oriented digital services such as social networking, e-commerce and online gaming, and the result is the burgeoning growth in hyperscale data centres.
“Enterprises are also choosing to house an ever-growing proportion of their data centre gear in colocation facilities, further reducing the need for on-premise data centre capacity. The rise of Generative AI technology and services will only exacerbate those trends over the next few years, as hyperscale operators are better positioned to run AI operations than most enterprises,” added Dinsdale.
Michael Winterson, Managing Director, European Data Centre Association (EUDCA); Anthony Milovantsev, Partner, Altman Solon; and Ivo Ivanov, CEO, DE-CIX, highlight how collaboration and tailoring services is a route to success for enterprises among dominant hyperscale activity.
Michael Winterson, Managing Director, European Data Centre Association (EUDCA)
In today’s fast-moving environment, with drivers such as Digital Transformation and the impact of AI, the CIO must make fundamental decisions around the proper deployment of infrastructure.
We’ve seen over the past 40 years – starting with mainframes, through the advent of the client-server space and up to cloud and the Internet – that technology is always more efficient at scale. Consequently, centralisation has been a fundamental feature of technology ever since. AI too, in the way it is currently designed, is fundamentally better as a mutualised service.
To a certain extent, CIOs will be compelled to centralise, but they will still have a multi-cloud environment. They will still have certain things that they consider sacrosanct to their business, personal information (PI), their data, their customer data, their private business operations, perhaps even a technology running on a factory floor. There may also be geographical issues for data for some territories that will require specific measures too.
Businesses can centralise, and still have the Edge Computing issue, in that certain things must live nearer to the business, the user, or the stakeholder. Therefore, businesses will be in a multi-cloud environment and may have a very complicated global network, a very complicated Zero Trust security system and will be constantly connecting a variety of technologies together and distributing them across their user base – be that local or global.
To achieve that, there will always be a role for the multi-tenant data centre to provide enterprises with services that don’t fit with the standard flavours of the hyperscale offerings. Every enterprise is consuming technology in the shape or form factor that fits their business, not the way the technology companies want to sell it.
We must stop assuming that we are going towards a monolithic cloud. Every enterprise that has tried to dedicate itself to one model of cloud has generally failed and is reversing as fast as it can several years later. A vanilla solution does not solve all problems.
Concerning the question then, it is not so much how to prepare for this proportion of market share, rather this is a rebalancing that is going on, but a rebalancing in a high period of growth. That means the whole pie is growing, and every slice is critical. No slice is likely to disappear but rather grow at different rates according to demand.
Businesses may reallocate where they invest, or their choice might be in a slower-growing section of the market, but still, the whole market is going to grow, with benefits for CIOs in terms of choice.
This is not a zero-sum game.
Anthony Milovantsev, Partner, Altman Solon
The rise of hyperscalers has created concerns in the data centre industry but has also opened opportunities for savvy smaller operators to retain – or even grow – their share of the more than US$300 billion annual global data centre market.
While smaller data centre operators do compete with hyperscalers in some cases, there can be value for both in working together in historically vexing areas like energy. The growth of hyperscaler services also enables smaller operators to create niche markets through specialised services and cater to environmentally conscious clients.
For non-hyperscale (retail) data centres and hyperscale-focused data centres, here are six strategic adaptations to the dominance of hyperscale services and own-build infrastructure:
- Power agreements with grid operators: Data centre operators can help hyperscalers with their time-to-market needs through advanced agreements on power commitments in areas with future build prospects. This would lower the wait time for a grid connection, which in some areas, including Northern Virginia, is approaching seven years.
- Specialisation and niche markets: Instead of competing head-on with hyperscalers, smaller operators can focus on niche markets where they can offer specialised services. This could include providing ultra-low latency services for financial sectors, compliance-focused storage for healthcare or government data, or localised data processing for IoT applications in smart cities.
- Edge Computing: With the rise of IoT and the need for real-time data processing, there’s a growing demand for Edge Computing solutions. Data centres can shift towards or incorporate Edge Computing, placing data processing closer to where it’s generated. This reduces latency, decreases bandwidth use and can be more efficient for certain applications, offering a service that hyperscalers might not provide as effectively from their centralised locations.
- Product expansion / Enhanced service offerings: Traditional data centres can transition from mere infrastructure providers to service providers by offering managed services, cybersecurity solutions, Disaster Recovery and bespoke IT consultancy. By doing so, they add value beyond what hyperscalers typically offer directly, focusing on customer service and tailored solutions.
- Sustainability initiatives: As environmental concerns grow, smaller data centres can differentiate themselves by focusing on green technology. While hyperscalers are often criticised for their large environmental footprint, smaller facilities can invest in renewable energy, advanced cooling technologies and sustainable practices to attract eco-conscious clients.
- Collaborative ecosystems: Building partnerships with hyperscalers instead of competing with them could be beneficial. This might involve becoming a part of a larger ecosystem where they handle overflow, specialised tasks, or act as regional partners for global giants, thereby gaining from the hyperscale trend rather than falling victim to it.
Ivo Ivanov, CEO, DE-CIX
Looking firstly at data centre operators, in particular colocation, we see the need for a mix of customers within their data centres. It may be tempting to lease these facilities to single hyperscalers – something which is happening with increasing frequency – but digital ecosystems depend on diversity.
Certainly, it can be tempting to pre-lease an entire facility to a single large-scale tenant, but many data centre operators will agree that this is not always an advantage in terms of margin, and it’s against diversity. You can’t build an ecosystem in isolation. Therefore, data centre operators need to aim for a healthy mix between single-tenant operations and multi-tenant facilities that enable network density, as well as AI-related builds.
The future will demand dense connectivity hubs for enterprise networks, similar to what we see in the carrier world today. Looking at the network presence of enterprises like Apple, Netflix, or Disney, their data centre requirements are no different from that of a carrier network presence.
Following the model of the carrier hotel, the future will demand more inclusive ‘network operator hotels’. In the enterprise world, this means enterprise-dense data centres where they can have their network Points-of-Presence (PoP).
The need for balance can also be seen in enterprise use of AI. Now, there is a strong demand for deployments for the training of LLMs and other large AI models, requiring specialised power densities. In current AI-related deployments, we see a ratio of around 70% high power density for training purposes compared to 30% for AI inference.
I have heard from multiple data centre operators the forecast of a significant shift in the coming years, in which this ratio will be reversed, so that 70% of AI-related deployments will be focused on inference and only 30% on training.
Given that AI inference does not demand equivalent power densities as training, this processing can be carried out in standard colocation facilities, taking some of the focus back away from the hyperscalers. There is a clear need for enterprise-focused data centres.
When we look at enterprises, the strength of the hyperscalers can be seen as being to their advantage, because the cloud players are heavily focused on serving enterprises. However, enterprises very often also want to operate their own infrastructure in hybrid scenarios, rather than placing everything in the cloud. Here, enterprise-rich colocation data centres again come into play.
Now, in some of the major business hubs, like New York City, colocation space is becoming scarce, so enterprises need to be creative about how to find the space they need.
The use of a distributed, data centre and carrier-neutral Internet Exchange offers two solutions to this challenge:
- On the one hand, the distributed interconnection model allows businesses to maintain efficient and secure data flows, even when their primary data centres are located outside of the main hub, as long as they choose a data centre that houses such a platform. This ensures they remain part of a vibrant digital ecosystem regardless of their physical data centre location.
- On the other hand, there is still space in these crowded hubs, but it can be that the available spaces are small and a bit piecemeal. Through distributed interconnection platforms, it is possible to connect these less ideal spaces with their current spaces as companies grow – providing seamless connectivity across multiple colocation data centre spaces, as well as with the larger ecosystem.
